Presentation on theme: "Causal-Comparative Research July 23, 2001 EAF 410."— Presentation transcript:
Causal-Comparative Research July 23, 2001 EAF 410
What is it? AssociationalAssociational Differences between/among groupsDifferences between/among groups –Categorical data –Differences already exist Ex post factoEx post facto Much research in medicine & sociologyMuch research in medicine & sociology
Examples What are the effects of teacher gender on student behaviorWhat are the effects of teacher gender on student behavior How are administrative decisions affected by genderHow are administrative decisions affected by gender What causes students to join the glee club? How are teaching styles affected by graduate studies?
Type 1 Exploration of effects caused by membership in a groupExploration of effects caused by membership in a group –What differences in student behavior are caused by gender?
Type 2 Studies causes of group membershipStudies causes of group membership –Why do ISU students enroll in EAF 410 during the summer session?
Type 3 Studies the consequences of an interventionStudies the consequences of an intervention –How do students who have taken EAF 410 evaluate research articles?
Causal Link Tentative Does gender cause differences?Does gender cause differences? –Girls read better –Boys are disciplined more Other probable links in the causal chainOther probable links in the causal chain
Basic Approach Noted difference between two groupsNoted difference between two groups Look for causes/consequencesLook for causes/consequences
Similarities with Correlational Research AssociationalAssociational Attempt to explain phenomenaAttempt to explain phenomena Often lead to experimental researchOften lead to experimental research Don’t allow variable manipulationDon’t allow variable manipulation Attempt to explore causationAttempt to explore causation
Causal-Comparative -compare groups -categorical variable(s) -crossbreak tables Correlational -individual scores -quantitative variables -scatterplots & r Similarities with Correlational Research
Steps Involved Identify & clarify research problemIdentify & clarify research problem –Several alternative hypotheses Select subjects (groups)Select subjects (groups) –Homogenous on some variables InstrumentationInstrumentation –No limits
Steps Involved (continued) DesignDesign –Groups differ on specific variable –Groups compared on other variable(s)
Evaluating Threats What other factors affect the comparison variable?What other factors affect the comparison variable? What is the likelihood of the groups differing on these factors?What is the likelihood of the groups differing on these factors? Evaluate threats on likelihood of their effect and plan to control.Evaluate threats on likelihood of their effect and plan to control.
Three Characteristics Information collected to describeInformation collected to describe Information collected by asking questionsInformation collected by asking questions Information collected from a sampleInformation collected from a sample
Major Purpose To describe the characteristics of a populationTo describe the characteristics of a population Not so concerned about whyNot so concerned about why Concerned about what isConcerned about what is
Types of Surveys Cross-sectional StudyCross-sectional Study –information collected at just one point in time Longitudinal StudyLongitudinal Study –information collected at different points in time
Types of Longitudinal Studies TrendTrend –Different samples –Population may change CohortCohort –Different samples –Population doesn’t change
Steps in Survey Research Clearly define objectivesClearly define objectives Construct clear questionsConstruct clear questions Sample (unit of analysis)Sample (unit of analysis) Collect dataCollect data –“live” to a group –mail –phone –face-to-face interviews
Preparing the Instrument Not too longNot too long Questions easy to answerQuestions easy to answer Carefully wordedCarefully worded –can be asked as written –will mean same to all –can be answered –will be answered
Suggestions for Closed- Ended Questions UnambiguousUnambiguous Simple focusSimple focus ShortShort Common languageCommon language “Unbiased” terms“Unbiased” terms Not leading questionsNot leading questions No double negativesNo double negatives
Suggestions for Questions Try them out on a similar sampleTry them out on a similar sample Make them look goodMake them look good Have follow-up questionsHave follow-up questions –contingency questions -- contingent on answer to previous question
Nonresponse Some will not return surveySome will not return survey Some items will not be answeredSome items will not be answered Can use random replacementCan use random replacement